This paper develops a consumption-oriented model of asset prices in a multigood economy that is, in principle, testable even when aggregate consumption of goods and their market prices are only partially observable. Previous studies show that, when there are m consumption goods, equilibrium expected excess returns on securities are functions of their covariances with m + 1 variables-aggregate consumption expenditure and market prices of consumption goods. Without making any further assumptions, the present model shows that a similar equilibrium relationship can be expressed in terms of covariances of asset returns with the following m + 1 variables: market prices of k consumption goods and aggregate consumption of m + 1 - k goods. Because the author's result provides researchers with some flexibility in choosing the set of m + 1 variables that measure riskiness of securities, it should lead to more powerful tests of the model. THIS PAPER DEVELOPS AN equilibrium model of asset prices in a multigood economy that is, in principle, testable even when aggregate consumption of goods and their market prices are only partially observable. If there are m consumption goods, then, according to previous studies (e.g., see Breeden [4]), the stochastic properties of a specific set of m + 1 variables must be estimated to test or to apply the consumption CAPM (CCAPM); i.e., the movement of aggregate consumption expenditure and market prices of m consumption goods must be observed. This paper, on the other hand, shows that the CCAPM can be empirically tested using any combination of m + 1 variables from the consumption opportunity set. In particular, if aggregate consumption of m - k goods can be observed, then market prices of only k + 1 goods have to be observed to test the model. To illustrate the results of this paper, we will consider three specific examples. First, if aggregate consumption of only one good can be observed, market prices of all goods will provide enough information to test or to apply the CCAPM. Second, if aggregate consumption of all goods can be observed, the market price of only one good has to be observed to test the model. Finally, we show that, when investors' direct utilities are separable, a smaller set of variables from the consumption opportunity set has to be observed to test the CCAPM.1